AI Paper Atlas

Research map / Vision-language and multimodal / Multimodal LLMs

HYDRA: A Hyper Agent for Dynamic Compositional Visual Reasoning

Fucai Ke, Zhixi Cai, Simindokht Jahangard, Weiqing Wang, Pari Delir Haghighi, Hamid Rezatofighi

ECCV 2024 · Vision-language and multimodal › Multimodal LLMs · approach: Reinforcement learning for visual reasoning · 19 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add RL agent controller to ViperGPT-style pipeline (ablation)ViperGPT · GQAaccuracy (%): 37.94 -> 43.71Improved
add RL agent on top of IR and sampling (ablation)HYDRA · GQAaccuracy (%): 46.93 -> 47.88Improved
add incremental reasoning (IR) (ablation)HYDRA · GQAaccuracy (%): 45.98 -> 47.88Improved
add multiple instruction sampling (ablation)HYDRA · GQAaccuracy (%): 47.07 -> 47.88Improved
sample multiple instructions without RL agent (ablation)ViperGPT · GQAaccuracy (%): 37.94 -> 39.84Improved
train RL agent on a different dataset (GQA instead of OK-VQA) (ablation)HYDRA · OK-VQAaccuracy (%): 48.63 -> 48.17No clear effect
train RL agent on OK-VQA instead of A-OKVQA (ablation)HYDRA · A-OKVQAaccuracy (%): 56.35 -> 55.94No clear effect
increase instruction samples to 10 (ablation)HYDRA RL agent · OK-VQAaccuracy (%)Hurt
use 5 instruction samples (ablation)HYDRA · OK-VQAaccuracy (%): 48.63Improved
replace BLIP2 with LLaVA1.5 (7B) as VFM in HYDRAHYDRA · GQAaccuracy (%): 47.9 -> 64.5Improved
wrap LLaVA1.5 (7B) in HYDRALLaVA1.5 (7B) · GQAaccuracy (%): 62.0 -> 64.5Improved
replace ViperGPT with HYDRA for visual groundingViperGPT · RefCOCO+IoU (%): 60.0 -> 61.1Mixed

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy